r - 在带有空值的 shiny 中使用 observeEvent

Question

我正在努力解决一个 react 性的 Shiny 问题，希望得到一些帮助。假设我有一个非常简单的应用程序，它显示给定一系列选择的平均高度:

library(tidyverse)
library(shiny)
library(DT)

df <- data.frame(cat = sample(LETTERS[1:4], 20, replace = T),
                 city = sample(c("Tokyo", "Madrid", "Paris"), 20, replace = T),
                 pop = sample(1:7, 20, replace = T),
                 altitude = sample(10000:15000, 20, replace = F))

ui <- fluidPage(
  selectizeInput("cat", "Category:",
                 choices = c("A", "B", "C", "D"),
                 multiple = T,
                 selected = "A"),
  selectizeInput("city","City:", 
                 choices = NULL, 
                 multiple = T),
  selectizeInput("pop","Population:", 
                 choices = NULL, 
                 multiple = T),
  dataTableOutput("table")
)

server <- function(input, output, session) {
  observeEvent(input$cat,{
    updateSelectInput(session,"city",
                      choices = df %>% filter(cat == input$cat) %>%
                        .$city %>% unique,
                      selected = "")
  })
  observeEvent(input$city,{
    updateSelectInput(session,"pop",
                      choices = df %>% filter(cat == input$cat & city == input$city) %>%
                        .$pop %>% unique,
                      selected = "")
  })
  output$table <- renderDataTable({
    df %>%
      filter(cat == input$cat, city == input$city, pop == input$pop %>% as.numeric) %>%
      summarise(mean_altitude = mean(altitude))
  })
}

# Run the application 
shinyApp(ui = ui, server = server)

我将如何修改此代码以允许在某些字段中进行空选择，并让其他选项对此完全 react ？现在，当您从上到下选择每个字段时，它是唯一的 react 。我希望它能够响应空值并且顺序无关紧要。

我可以在查询时使用复杂的案例来做到这一点，但我的实际数据集有大约 10 个不同的输入，并且通过扩展大量不同的组合，所以这样做来处理每一个可能的组合是不可行的。

Answer 1

observeEvent(input$id, {
    code 
    ...
}, ignoreNULL = FALSE)



?observeEvent

这对我来说也是最近的发现